Refactor to fix all but the Id -> ID lint errors

In order to support more contributors, prepare the project for clean
lint passes.

The lint error for Id in names will produce breaking changes. If they
were fixed only non-breaking changes, it would create inconsistency.
They are an unfortunate relic from this library coming before that lint
rule. Consistency and interface comptability trumps the `Id` rule.

auth.go

@@ -22,12 +22,14 @@ type PlainAuth struct {
 	Password string
 }
 
-func (me *PlainAuth) Mechanism() string {
+// Mechanism returns "PLAIN"
+func (auth *PlainAuth) Mechanism() string {
 	return "PLAIN"
 }
 
-func (me *PlainAuth) Response() string {
-	return fmt.Sprintf("\000%s\000%s", me.Username, me.Password)
+// Response returns the null character delimited encoding for the SASL PLAIN Mechanism.
+func (auth *PlainAuth) Response() string {
+	return fmt.Sprintf("\000%s\000%s", auth.Username, auth.Password)
 }
 
 // Finds the first mechanism preferred by the client that the server supports.